Difficulty incorporating the paretic arm and hand into everyday activities is a significant contributor to reduced function and quality of life after stroe. While behavioral interventions can improve arm motor function after stroke, outcomes are often incomplete, with a large amount of variability in response to treatment. One important but not yet explored factor that may influence recovery is the ability to effectively and efficiently engag a set of motor regions (or motor circuit) based on changing task demands. Our long-term goal is to examine the plasticity of individual motor circuits after stroke and determine whether targeting these circuits during training benefits function. As a first step, the premotor-prefrontal component of the motor network is a prime target for motor practice given the significant role of these regions in complex movements and motor recovery after stroke and their well-documented role in motor action selection. The proposed study will determine the ability to target the premotor-prefrontal motor circuit over practice through systematic changes in task conditions. Individuals post-stroke will practice a motor task that requires action selection to determine movement response over four days (change in reaction time). Brain imaging will be performed at the beginning and at the end of practice; both magnitude of activation and functional connectivity between regions will be used to quantify brain activation. The specific aims of this project are to: 1) determine the effect of a period of practice targeted at the premotor-prefrontal motor circuit through the addition of action selection demands to movement; and 2) determine whether behavioral improvement over practice can be predicted by initial performance. We hypothesize that as behavioral performance improves, neural activation in premotor and prefrontal regions in both the ipsilesional and contralesional hemisphere will become more efficient, as demonstrated by decreased magnitude of activation and increased connectivity between regions. We also hypothesize that individuals who are more efficient at engaging the premotor-prefrontal motor circuit during motor action selection on day one (less activation magnitude and greater premotor-prefrontal connectivity) will show greater behavioral improvement over practice. The approach is innovative in that it incorporates a change in practice condition to drive changes in a specific component of the motor circuit after stroke and includes measures of brain activation magnitude and connectivity between brain regions which provide related but unique data on neural activation. The proposed project is significant in that i provides an important first step in determining whether targeted engagement of a specific component of the motor circuit during practice is feasible after stroke. The results from this project will set the stage for future work aimed at the development of an intervention protocol that targets this circuit or other circuits for the purpose of improving arm function and quality o life in individuals with motor dysfunction due to stroke.